基于TOPMODEL的中小河流分布式洪水预报模型及其应用

近年来广东省中小河流洪水灾害频发,研究中小流域洪水预报模型对预防洪水灾害具有重要意义。本研究改进了传统的TOPMODEL模型,以栅格为单元构建了基于TOPMODEL的中小河流分布式洪水预报模型,并以高田水上游流域为研究对象,采用SCE-UA优选算法对1975—2008年的8场洪水进行参数优选,并对2010—2012年的4场洪水进行模型验证。模拟结果表明,本次预报在率定期和验证期平均确定性系数为0.85,径流深预报合格率为92%,峰现时间预报合格率为100%,模型模拟效果较为理想。用改进后的模型研究了某场洪水汇流过程不同时刻的空间分布情况,表明改进后的TOPMODEL模型能够直观地反映洪水的时空变化,模型具有一定的合理性和实用性。

Application of TOPMODEL-based distributed flood forecasting model for small-sized and medium-sized rivers

In recent years，flood disasters have occurred high-frequently in small-sized and medium-sized rivers of Guangdong Province． It is of great practical significance to study flood forecasting model in preventing flood disasters in small-and mediumsized basins． In this study we improve traditional TOPMODEL，using the grid as unit to construct a TOPMODEL-based distributed flood forecasting model for small-and medium-sized rivers． SCE-UA optimization algorithm is applied to calibrate the model parameters for eight flood events from 1975 to 2008 in Gaotian-Shui river upstream basin． Four flood events from 2010 to 2012 are adopted to verify the model． The simulation results show that qualified rate of runoff depth forecasting and peak timing forecasting are 92% and 100% respectively in calibration and verification period． Average deterministic coefficient of this forecast reaches 0. 85． Therefore，the improved model performs well as applied in the study catchment． The improved model is used to study the spatial distribution of a flood confluence process at different time． It shows that the improved TOPMODEL can intuitively reflect the temporal and spatial variation of flood and indicates that the improved model has certain rationality and practicability．